A case–control study analyzing mannitol dosing for prevention of cisplatin

Mannitol is an osmotic diuretic given routinely as part of cisplatin regimens to prevent nephrotoxicity, but there are limited data on the ideal dosage. At our center, three different doses of mannitol are used: 12, 20, and 40？g per cycle for cisplatin doses of ≥50？mg/m2. The primary objective was to determine if variations in mannitol dosing significantly influence the incidence of cisplatin-induced acute nephrotoxicity. A case–control study was performed. Electronic records of 1462 consecutive outpatients who received cisplatin at ≥？50？mg/m2 per cycle between January 2010 and December 2014 were reviewed. Patients experiencing nephrotoxicity of any grade within 30 days of last cisplatin dose, as defined by NCI CTCAE 4.0, were matched to a minimum of two and maximum of five controls based on the following criteria: age ±？5 years, baseline estimated glomerular filtration rate ±？10？ml/min/1.73？m2, cisplatin dose per cycle, and presence of diabetes. Conditional logistic regression was used to identify baseline predictors of cisplatin-induced acute nephrotoxicity. Of the 1245 included patients, 237 had nephrotoxicity and 1008 were matched controls. Median baseline estimated glomerular filtration rate for cases and controls were 83 and 80？ml/min/1.73？m2, respectively. A total of 3.8% of cases experienced ≥？grade 3 nephrotoxicity. Univariable analysis showed that diabetes, lymphoma, low baseline estimated glomerular filtration rate, and low baseline magnesium level were significantly associated with nephrotoxicity, whereas mannitol dosing did not show any association (odds ratio 1.08; p？=？0.29). In multivariable analysis, diabetes and lymphoma retained statistical significance, but baseline estimated glomerular filtration rate and baseline magnesium level showed nonsignificant associations with nephrotoxicity. Cisplatin-induced acute nephrotoxicity remains common in patients with good baseline renal function despite preventive measures. Diabetes and lymphoma are predictors of nephrotoxicity, whereas mannitol dosing has no significant influence, suggesting that doses may be standardized across cisplatin regimens